Development of Chronic Sphenoid Sinusitis After Sellar Reconstruction with Medpor Porous Polyethylene Implant

Abstract

■ OBJECTIVE:

The Medpor porous polyethylene implant is reported to be safe and effective for sellar reconstruction after transsphenoidal surgery (TSS). However, we have observed several cases of delayed chronic sphenoid sinusitis related to the implant. The purpose of this study is to describe the presentation and management of implant-related sphenoid sinusitis after sellar reconstruction.

■ METHODS:

This is a retrospective study of patients who underwent endonasal TSS with Medpor sellar reconstruction between December 2008 and January 2013 at a tertiary care institution. Patient demographics, initial surgical management, sinonasal symptoms, postoperative imaging, sinusitis management, and resulting outcomes were analyzed.

■ RESULTS:

From 2008—2013, 139 patients underwent sellar reconstruction using Medpor. Five patients (3.6%) presented between 8 and 60 months after surgery with chronic sphenoid sinusitis that required surgical management. All 5 patients presented as outpatients for management of headaches and nasal drainage, 4 patients experienced chronic nasal congestion, and 3 patients noted recurrent sinusitis. At the time of revision surgery, all 5 patients were found to have mucosal inflammation and edema surrounding the implant, and 4 of the 5 had an exposed or partially extruded implant that was removed.

■ CONCLUSIONS:

Reconstruction of the sellar floor may be performed after TSS to prevent postoperative complications. Although porous polyethylene implants have previously been described as safe and effective for this purpose, surgeons should be aware of the risk of subsequent implant extrusion and chronic sphenoid sinusitis that can occur in a delayed manner.

INTRODUCTION

Tumors of the sellar region account for approximately 9.1%—15.3% of intracranial tumors.^1,2 Transsphenoidal surgery (TSS), which allows access to the sella as well as the parasellar region via extended approaches, is the mainstay of treatment for patients with sellar or parasellar lesions.

Cerebrospinal fluid (CSF) leak is still a major complication of TSS, with rates reported in the literature ranging from 2%—13.8%.^3–8 It has been common practice to reconstruct the sellar floor to minimize the risk of delayed CSF leak and secondary empty sella after TSS. In some surgical practices, this is done routinely when possible. There have been a wide variety of techniques proposed for sellar reconstruction, including both autologous and synthetic biomaterials.^9–13 Autologous materials such as septal or sphenoid bone, cartilage, fascia, fat, and muscle have traditionally been preferred because of their biocompatibility. However, biomaterials such as porous polyethylene, titanium, or silicone plates have been used to obviate the need for additional surgical site morbidity and prolonged operative time.

The Medpor implant (Stryker, Kalamazoo, Michigan, USA) is a commercially available non-absorbable porous polyethylene implant used to reconstruct many surgical defects and has been used to reconstruct the sellar floor. Two studies have been published supporting the safety and efficacy of the Medpor implant.^9,14 However, in our practice, we have noted a series of patients who developed chronic sphenoid sinusitis, ultimately requiring revision sinus surgery and, in most cases, implant removal. The purpose of this study was to examine the incidence and features of patients who developed chronic sphenoid sinusitis after porous polyethylene (Medpor) reconstruction of the sellar floor in a large patient group.

METHODS

A retrospective chart review was performed on all cases of microscopic or endoscopic TSS performed by a single surgeon between January 1, 2008 and December 31, 2013, to determine cases in which the Medpor porous polyethylene implant was used for sellar reconstruction. Cases in which other alloplastic or autologous implants were used have been excluded from this review. In addition, cases in which the Medpor implant was removed during revision TSS were only analyzed until the time of implant removal. All cases were performed at the University of Colorado Anschutz Medical Campus, and data were extracted from charts and examined with the approval of the Colorado Multiple Institutional Review Board (COMIRB #17–2387). Patient demographics, characteristics, tumor histology, intraoperative reconstruction methods, perioperative complications, post-operative symptom and treatment course, and long-term complications were examined.

RESULTS

Medical charts were queried for all patients >18 years old who underwent transsphenoidal pituitary surgery between January 1, 2008 and December 31, 2013 by a single surgeon (K.O.L.). A total of 139 patients of these cases underwent sellar reconstruction using a Medpor implant trimmed to the defect size at the end of the procedure. The treating surgeon used the implant in a routine fashion during this time period for reconstruction of the sellar floor when septal bone remnants were of insufficient quality. All patients received 48—72 hours of perioperative antibiotics while in the hospital. Patient demographics are summarized in Table 1. The patient population ranged in age from 19—85 years, with a median age of 54, and had near-equal distribution of men and women. Most patients who underwent sellar reconstruction using Medpor had never smoked or were former smokers, 61.2% and 26.4%, respectively.

Table 1.

Demographics and Patient Characteristics of Medpor Implant Recipients

	Patients with Medpor Implants	Patients with Medpor Implants Undergoing Surgical Revision for Chronic Sphenoid Sinusitis
Patients
Total number	139	5
Age, median years (range)	54 (19–85)	58 (29–70)
Sex, n (%)
Male	70 (50.4)	3 (60)
Female	69 (49.5)	2 (40)
Smoking status, n (%)
Current	18 (13.4)	1 (20)
Former	34 (25.4)	3 (60)
Never	82 (61.2)	1 (20)
Prior TSS, n (%)
Yes	43 (30.9)	2 (40)
No	96 (69.1)	3 (60)
Prior radiation, n (%)
Yes	2 (1.4)	1 (20)
No	137 (98.6)	4 (80)

TSS, transsphenoidal surgery.

Table 2 summaries the pathology, size, and extrasellar involvement of the patients who underwent Medpor cranioplasty. Of the 139 identified cases, 83% were pituitary adenomas— with true null cell adenomas and gonadotroph adenomas being the most common—comprising 25.2% and 19.4% of sellar pathologies, respectively. Most lesions were >1 cm in size and 70.5% demonstrated extrasellar extension. In most patients, the implant was used for reconstruction on completion of primary surgical management, with only 30.9% of patients presenting after prior TSS. Only 1.4% of patients had previously received radiation therapy for their sellar pathology, and only 1.1% of our patient population received postoperative adjuvant radiation.

Table 2.

Sellar Lesion Characteristics of Medpor Implant Recipients

	Patients with Medpor Implants	Patients with Medpor Implants Undergoing Surgical Revision for Chronic Sphenoid Sinusitis
Final pathology, n (%)
Non-secretory adenoma	35 (25.2)	2 (40)
Pituitary adenoma, polyhormonal	18 (12.9)	0 (0)
Gonadotroph adenoma	27 (19.4)	0 (0)
Corticotroph adenoma	16 (11.5)	1 (20)
Prolactinoma	12 (8.6)	0 (0)
Somatotroph adenoma	7 (5.0)	0 (0)
Rathke cleft cyst	4 (2.9)	0 (0)
No lesion or unable to determine	13 (9.4)	1 (20)
Other	7 (5.0)	1 (20)
Longest dimension, n (%)
<1.1 cm	10 (7.2)	0 (0)
1.1–2.0 cm	23 (16.5)	3 (60)
2.1—3.0 cm	26 (18.7)	0 (0)
>3.1 cm	16 (11.5)	0 (0)
Unknown	64 (46.0)	2 (40)
Extrasellar extension, n (%)
Yes	98 (70.5)	4 (80)
No	36 (25.9)	1 (20)
Unknown	5 (3.6)	0 (0)

Most patients in this cohort (97.1%) underwent microscopic TSS; the remaining 2.9% underwent endoscopic TSS. In the overall cohort, 79.0% were noted to have a gross total resection, and 45.3% of patients were noted to have an intraoperative CSF leak. In the 139 subjects who underwent Medpor reconstruction, the majority of the sellar defects were reconstructed with either Medpor alone (52.5%) or in combination with a dural sealant, such as fibrin glue or Duraseal (Integra Life Sciences Corp., Plainsboro, New Jersey, USA) (37.4%). The remaining patients’ reconstruction also used fat, fascia lata, or duragen in addition as either an underlay or overlay graft. Table 3 summarizes the reconstructions used.

Table 3.

Sellar Reconstructions

	Patients with Medpor Implants	Patients with Medpor Implants Undergoing Surgical Revision for Chronic Sphenoid Sinusitis
Selllar reconstruction, n (%)
Medpor alone	73 (52.5)	2 (40)
Medpor and dural sealant	52 (37.4)	2 (40)
Other (combination of Medpor, fat, fascia, duragen, dural sealant)	14 (10.1)	1 (20)

Patients were followed for a median of 26 months after surgery (range 0—105 months). Nine patients were specifically referred for evaluation by the rhinology service, with 6 complaints of chronic sinusitis, 2 concerns for Medpor plate extrusion based on magnetic resonance imaging findings, and 1 complaint of persistent postnasal drip. The median time to rhinology evaluation was 22 months after TSS (range 4—60 months).

On rhinologic evaluation, the most common patient complaint was rhinorrhea, which was noted in 89% of presenting patients. Nasal congestion, headache, and reports of recurrent sinus infections were also common, with 78%, 67%, and 44% of patients noting those symptoms, respectively. Initial standard of care sinus therapies were instituted according to the American Academy of Otolaryngology-Head and Neck Surgery Sinusitis guidelines. Of the 9 patients evaluated by rhinology, 4 were successfully managed conservatively with observation, nasal saline solution irrigations, or with saline solution rinses combined with nasal steroids. One of these patients had a long history of chronic sinusitis previously managed with endoscopic sinus surgery and presented with recurrent symptoms of thick postnasal drip. On nasal endoscopy and imaging, he was found to have mild recurrence of his nasal polyposis within the ethmoid sinuses. He has since been followed by rhinology and managed successfully with nasal steroid irrigations. Two patients presented for evaluation of nasal obstruction with reports of nasal drainage. In addition, routine surveillance magnetic resonance imaging demonstrated sphenoid mucosal thickening on both patients. However, nasal endoscopy did not identify active sinonasal disease or evidence of plate exposure. One patient was noted to have nasal obstruction owing to a septal deviation and the other was diagnosed with rhinitis. The last patient presented with fatigue and sinus headaches. Imaging and nasal endoscopy ruled out sinus etiology, and the patient was diagnosed with a primary headache disorder.

Despite medical intervention, 5 patients (3.6%) ultimately required surgical revision (revision sphenoidotomy with debridement). Patient demographics are summarized in Table 1. The revision surgical group comprised 2 women and 3 men, with a median age of 58 (range 29—70) years old. Surgery took place 8—60 months after initial TSS. Two patients had undergone multiple TSS and 3 had only undergone a single prior TSS. Only 1 patient had received radiation therapy to the sella prior to rhinology evaluation. Sellar lesion characteristics are summarized in Table 2. There were no consistent tumor characteristics among patients who required revision surgery: 3 patients had pituitary adenomas, 1 patient did not have any identifiable lesion after exploration, and 1 patient was noted to have Langerhans cell histiocytosis involving the sellar and parasellar region. Of the 3 adenomas managed, all were 1—2 cm in the largest dimension and had extrasellar extension. Three patients were noted to have had intraoperative CSF leaks at the time of implant placement. At the time of revision surgery, all 5 patients were noted to have headaches and rhinorrhea; however, none had CSF rhinorrhea. Four patients also complained of congestion and 3 reported recurrent sinusitis. Figure 1 demonstrates the range of preoperative sinus computed tomography findings, ranging from frothy secretions to complete sphenoid opacification. Because of their radiolucency, porous polyethylene implants are not visualized on imaging. Similarly, there was a wide range of intraoperative findings including purulence, crusting, fungal debris, scarring, and polypoid change. Four of the 5 patients were found to have exposed or partially extruded plates at the time of surgery, whereas the fifth patient was found to have significant mucosal inflammation overlying a fully mucosalized implant. Figure 2 demonstrates representative intraoperative findings from 2 patients. Intraoperative cultures demonstrated 2 patients with chronic Pseudomonas aeruginosa and 1 patient with a mix of Staphylococcus aureus and Streptococcus pneumoniae.

Figure 1.

Preoperative coronal computed tomography imaging findings from 3 subjects undergoing Medpor explantation. (A) Frothy secretions within the sphenoid; (B) rounded density consistent with early sphenoid mucocele or polypoid inflammation; (C) complete sphenoid opacification with associated neo-osteogenesis.

Figure 2.

Intraoperative endoscopic findings of 2 separate patients. (A) Crusting within the central sphenoid sinus (left), after debridement shows a partially extruded Medpor plate (right). Cultures demonstrated Pseudomonas aeruginosa. (B) Purulence and edema encountered on entry to the sphenoid sinus (left), with partial exposure of the porous polyethylene implant in the midline sella (right).

After surgical management, all patients did well overall. Four of the 5 patients had resolution of their nasal drainage. One had significant improvement of nasal drainage, but continued to have some symptoms, and was managed with ipratropium nasal spray. Three of the 5 patients noted resolution of their headaches in the early postoperative period. One patient continued to note persistent similar headaches, whereas another developed a new type of headache, more consistent with migraine, and were recommended for headache management by the neurology service. All patients who noted nasal congestion and recurrent sinusitis noted improvement in their symptoms postoperatively.

DISCUSSION

In many surgical practices, including this current series, reconstruction of the sellar defect at the end of pituitary surgery is done routinely. The most commonly cited true indication for sellar reconstruction, beyond surgeon preference, is the presence of a CSF leak.^4,15–17 Although not as critical, other indications for reconstruction include allowing for more rapid mucosal healing and earlier resumption of mucociliary function, prevention of arachnoid herniation with delayed CSF leak, and prevention of optic nerve prolapse into a large defect.^9,14,18 In the current series, an intraoperative CSF leak was noted in 45.3% of patients who underwent sellar reconstruction with Medpor plates.

During sellar reconstruction, the ideal option would be to use autologous tissue, because it is completely biocompatible.^11,19 However, the use of autologous material may result in donor site morbidity, difficulties in contouring the available material, and potentially increased operative time.²⁰ In addition, there is a limit to the availability of autologous implants, particularly in revision surgery, and if the surgeon is seeking a rigid reconstructive layer.²¹ In light of those limitations, various alloplastic materials have thus been used as a viable alternative to autologous material. A wide range of alloplastic materials have been used in TSS, such as ceramic, silicone, titanium, and porous polyethylene.^10,12,13,21

The Medpor implant is a non-absorbable porous polyethylene implant that has been used in craniofacial reconstruction since the 1990s. It is available in many different sizes and shapes, allowing it to be easily contoured to accommodate various defects. Its application has been described in reconstruction of the medial and lateral orbital walls, rhinoplasty, midface and mandibular augmentation, and cranioplasty.²⁰ Porous polyethylene implants theoretically may be favorable in that its high-density pores have been shown to induce rapid fibrovascular and soft tissue ingrowth, which allow for incorporation into the surrounding tissue and increased implant stability. This fibrovascular tissue ingrowth is believed to allow cellular products to permeate the implant early on, ameliorating the risk of infection.^14,19,22 In addition, the rapid soft tissue integration of the implant allows for increased stabilization rate, reducing the risk of implant migration and extrusion.^19,23 Other potential benefits of a porous polyethylene implant are that it is highly inert and does not induce significant foreign body reaction in the surrounding tissue.^9,20 Major disadvantages of these implants are their radiolucency, which limits its evaluation on postoperative imaging, its permanency and its inherent avascular nature that is at risk when exposed to microbial colonization in the sinuses.²⁴

The Medpor implant was first described for skull base cranioplasty by Park and Guthikonda in 2004.⁹ In their series of 10 patients followed for 9—23 months, they did not note any infectious or implant related complications. Liebelt et al.¹⁴ further discussed the safety of the Medpor implant for sellar reconstruction in their review of 136 consecutive implants that were followed for at least 1 year after TSS. In their follow-up period, they did not have any cases of implant extrusion, however, this may not have been systematically evaluated. They reported that 6 patients developed “sinus irritation and/or drainage,” all of which were successfully managed with conservative therapy and did not require revision surgery. In our case series, a total of 9 patients were evaluated for potential plate related complications, 5 of which required revision endoscopic sinus surgery for chronic sphenoid sinusitis (3.6%). In addition, at the time of surgery, 4 patients were noted to have non-mucosalized or partially extruded plates, a feature that was not noted in Liebelt’s study. Since the time of this study, our practice has converted from the use of porous polyethylene to resorbable plates, when needed. Additionally, further scrutiny is applied in the decision-making of when the sellar floor requires reconstruction in the absence of CSF leak, and surgeon preference appears to play a key role in the absence of any guidelines to direct this intraoperative decision.

A primary concern of alloplastic implantation is the development of biofilms, which are specialized bacterial colonies that are extremely difficult to eradicate even when appropriate antibiotics are administered. Biofilms can form on any type of surface; however, there is a reasonable concern that the porous polyethylene implant, with its irregular surface and porous nature, may serve as an ideal structure for biofilm formation. In fact, guinea pig studies of porous polyethylene implants demonstrated biofilm formation within 7 days of implantation.²⁵ When placed adjacent to the sinus, which even in the healthy state may harbor many biofilm-forming respiratory pathogens, an environment may exist where colonization of the implant and biofilm formation can outcompete vascular ingrowth and tissue coverage.²⁶

The main limitation of this study is that many patients did not follow up beyond the immediate postoperative period. In addition, those that did follow up were monitored by the neurosurgical and endocrinology services, who may not have been systematically assessing patients for signs of chronic rhinosinusitis. The retrospective nature of this cohort may also not identify all affected patients. As such, our study possibly underrepresents the actual population of patients who develop chronic sphenoid sinusitis after sellar reconstruction using porous polyethylene implants. We did not identify any specific risk factors for implant extrusion or infection, but advocate that its use should be carefully considered, and when used patients should be aware of potential for delayed postoperative sinus symptoms.

CONCLUSIONS

After TSS, sellar reconstruction is often performed either for management of intraoperative CSF leak or for the prevention of postoperative complication. Although porous polyethylene implants have previously been considered a safe and effective option for sellar reconstruction, their use may be associated with delayed development of chronic sphenoid sinusitis and plate extrusion. As such, neurosurgeons may consider using alternative reconstructive options, or patients should be monitored for the development of sinonasal symptoms and referred to otolaryngology if any concerning symptoms develop.

Abbreviations and Acronyms

CSF

Cerebrospinal fluid

TSS

Transsphenoidal surgery